An estimated 65% of new nursing home admissions suffer from dementia, and concern for the quality of their care is growing. In response, the use of special care units (SCUs) for persons with dementia is proliferating. The growth of SCUs has been controversial, because it has proceeded in an unregulated fashion and without clear benefit for residents. Health care professionals recognize the need to examine the particular features of long-term care which are important in the care of nursing home residents with dementia. Given the limited research in this area, the proposed study is designed to: (l) describe and compare the characteristics of a diverse set of 59 nursing homes (including 14 SCUs) which contain a mix of patients similar to those in U.S. facilities; and (2) assess the relationship between facility characteristics (e.g. staff/resident ratio; resident involvement in activities) and health outcomes (mortality, morbidity, health care use) for demented residents during the year following admission to the nursing home. Secondarily, it will examine whether the effect of facility characteristics influence functional outcomes (cognition; independence in activities of daily living; behavior) for survivors at one year post- admission, and whether the facility characteristics which are beneficial for demented residents are also of benefit for non-demented residents. This study will build on and use data from a current study of new admissions to Maryland nursing homes (ROl AG082II). All individuals (N=2000) aged 65 or older admitted for the first time to a 24% stratified random sample of 53 nursing homes during one year are being enrolled. Within one month of admission, evaluators interview residents, family and staff and abstract charts to obtain information about medical, cognitive and functional status. An expert panel of neurologists and psychiatrists is examining these data and making a determination of dementia following DSM-III-R criteria. Mortality, morbidity (infections, pressure sores, falls, febrile episodes) and nursing home and acute care use are being monitored through one year by chart review. The proposed project will use data collected in the current study and will add 6 SCUs (and 60 new admissions) to the 53 facilities. Detailed information about the facility (the treatment setting and the type of care provided) will be obtained from direct observation, interview with administrative personnel, and review of care plans. Also, information on functioning will be obtained for those surviving to one year. This study is unique in its examination of the relationship of facility characteristics to multiple quality of life outcomes for a large cohort of residents who are representative of all U.S. admissions and whose diagnosis of dementia is determined by an expert panel.